Progress in FY 2002 has been in the following areas: 1. Methodology. We have developed new solid state NMR techniques that allow measurements of polypeptide backbone and sidechain torsion angles in samples that contain uniformly 13C- and 15N-labeled residues at specific sites, either isolated residues or contiguous stretches of residues. The techniques depend on new radiofrequency pulse sequences for recoupling anisotropic chemical shifts and dipole-dipole couplings under magic angle spinning conditions. So far, these techniques have been demonstrated on polycrystalline model compounds. Initial applications in structural studies of amyloid fibrils and an HIV-1 related peptide/antibody complex are under way. In a separate methodological development, we have developed the first techniques for efficient excitation of high-order multiple quantum coherences in solid state NMR under high-speed magic angle spinning. We have previously used multiple quantum spectroscopy of nonspinning samples to constrain the structures of amyloid fibrils. The new techniques extend the range of application of multiple quantum spectroscopy significantly. 2. Amyloid fibril structure. We have developed the first truly experimentally based structural model for amyloid fibrils formed by the Alzheimer's beta-amyloid peptide. The model is based on solid state NMR measurements of intermolecular distances, NMR chemical shifts that constrain the peptide backbone conformation, torsion angle measurements, and measurements of fibril dimensions and mass-per-length by electron microscopy. In this model, the peptide adopts a conformation with a 10-residue disordered N-terminus and two structurally ordered beta-strand segments separated by a "bend" segment. The beta-strands form intermolecular beta-sheets with a parallel, in-register alignment. Fibrils with minimal dimensions and mass (called protofilaments) consist of two molecular layers joined by hydrophobic interactions. This model appears to be stabilized by both favorable electrostatic and favorable hydrophobic energies. 3. Studies of amyloid fibril growth. In typical preparations, amyloid fibrils formed by a single peptide or protein display a variety of morphologies in electron microscope images. Solid state NMR spectra of such samples typically exhibit multiple NMR signals for single 13C or 15N sites. The connection between the NMR spectra and the morphologies has been unclear. Recently we have shown that subtle changes in fibril formation conditions lead to distinct fibril morphologies that are associated with specific features in the NMR measurements (i.e., specific chemical shifts and specific nonsequential 13C-15N contacts). Thus, we have shown that a given peptide or protein can adopt different, distinct molecular structures in amyloid fibrils, i.e., the sequence does not determine the molecular structure uniquely. Our ability to control fibril morphology and molecular structure by controlling fibril formation conditions will facilitate further proof and refinement of our structural model for Alzheimer's beta-amyloid fibrils. 4. Studies of an HIV-1 related peptide/antibody complex. We have initiated a new project to determine the bound conformation of a peptide derived from the V3 loop of HIV-1 gp120 in complexes with a broadly neutralizing human monoclonal anti-gp120 antibody (Fv fragements). In this project, the peptide is uniformly 13C- and 15N-labeled at seven residues in the epitope region, allowing conformational measurements by our newly-developed methods appropriate for samples with uniformly labeled residues. Because the antibody (447-52D) is broadly neutralizing, it will be interesting to examine conformational differences between V3 peptides from different HIV-1 strains, which have significant sequence differences. This project will also demonstrate the general feasibility of solid state NMR studies of peptide/protein complexes, setting the stage for studies of complexes between peptides and integral membrane receptor proteins.